Researchers collaborate on Army artificial intelligence/machine learning and autonomy efforts
ADELPHI, Md. -- U.S. Army researchers are collaborating with University of Missouri-Columbia scientists to develop state-of-the-art capabilities for Soldiers.
Specifically, the researchers are focused on machine learning capabilities and addressing challenges in edge computing for drone video analytics, object detection and tracking, automatic landmark recognition and autonomous unmanned aerial vehicle-borne sensor configuration.
The U.S. Army Combat Capabilities Development Command, now referred to as DEVCOM, Army Research Laboratory enabled this collaboration through a $4.175 million cooperative agreement.
The agreement between the laboratory and the University of Missouri on the Video Analytics and Image Processing for Multiview Scene Understanding project began in September 2018.
After the successful annual review in August 2020, the Army extended the project performance through March 2022.
This collaboration supports and helps advance ARL’s artificial intelligence/machine learning and autonomy efforts, and is aligned more broadly with the Army’s Modernization Priorities such as Next Generation Combat Vehicle, Network and Soldier Lethality.
In support of this project, university Electrical Engineering and Computer Science professor Dr. K. Palaniappan leads a team of faculty members, postdoctoral researchers and graduate and undergraduate students with expertise in visual target tracking, geo-localization, drone based wide-area motion imaging, or WAMI, and high-speed computing on size, weight and power, or SWaP, constrained platforms.
“Dr. Palaniappan and his team have been historically strong in National Institutes of Health biomedical and NASA remote sensing projects, and I invited him to extend their work in variational level set image segmentation methods and spatiotemporal flux tensor moving object detection to analyze aerial WAMI datasets, a focused challenge at the Air Force Research Laboratory in 2007,” said Dr. Guna Seetharaman, who is now the Navy senior scientist for Advanced Computing Concepts and chief scientist of Computation at the Naval Research Laboratory and member of the advisory board for the ARL-MU project. “This evolved into a significant component of the AFRL C4ISR Enterprise to the Edge, or CETE program, and a data collection and testing exercise at Fort Leonard Wood, Missouri. It is exemplary to see that this prior work is being transitioned to meet new U.S. Army Futures Command requirements.”
According to Dr. Raghuveer Rao, chief of the laboratory’s Intelligent Perception Branch and cooperative agreement manager for this project, the main technical objectives of the project are the development of machine-learning approaches to exploit the richer information content provided by multi-view observations of an area of interest using both ground and air-based platforms, in conjunction with prior foundational data, for obtaining improved situational awareness for Soldiers.
The UM lab’s foundational expertise in computer vision and video analytics enabled them to create the first end-to-end processing chain for wide area motion imagery with pivotal support from Leonard Wood Institute, which catalyzed the current ARL project, Palaniappan said.
“This partnership allows us to accelerate emerging AI technologies to rapidly process large volumes of multi-facet video streams for enhanced scene perception, tactical situational awareness, geospatial context and advanced mobility at city scale, in degraded and denied environments,” he said. “We’re grateful for the opportunity to continue the ARL collaboration and are excited to contribute to emerging military and civilian drone applications and national testbeds.”
Machine-learning solutions developed as part of this effort for automatic tracking of dismounts from unmanned aerial vehicles have formed part of many high-level demonstrations to Army senior leadership, Rao said.
“It was incredible to see the results of this collaborative partnership after just one year,” said Dr. Mark Tschopp, regional lead for ARL Central in Illinois. “ARL’s mission is to take basic research and operationalize this science for transformational overmatch for the Army. Part of this is expanding the team to include experts in academia, small businesses and industry to push concepts and ideas into future capabilities for the Army.”
The University of Missouri is bringing subject matter experts, unique algorithms and capabilities, and great student researchers in video analytics into a partnership with the Army to develop the next generation of talented researchers, to advance these technologies, and to provide future capabilities for the warfighter, Tschopp said.
“It is our privilege to support our U.S. National Defense,” said Noah Manring, interim dean of Mizzou Engineering. “Our faculty and students are among the best in the world for applying artificial intelligence and machine-learning to real world problems; and here we are using these methods to keep our Soldiers and pilots out of harm’s way during combat. I commend Prof. Palaniappan and his team for this work, and appreciate the Army’s trust in our abilities to lead such an important effort.”
CCDC Army Research Laboratory is an element of the U.S. Army Combat Capabilities Development Command. As the Army’s corporate research laboratory, ARL is operationalizing science to achieve transformational overmatch. Through collaboration across the command’s core technical competencies, CCDC leads in the discovery, development and delivery of the technology-based capabilities required to make Soldiers more successful at winning the nation’s wars and come home safely. CCDC is a major subordinate command of the Army Futures Command.